This Project focuses on readers' sensitivity to morphological structure in written words, and it compares that sensitivity across readers of languages representing a variety of morphological and orthographic types. Morphological relationships impart structure to the otherwise arbitrary mapping between word forms (orthographic and phonological) and word meanings, and that structure may play a role in word recognition. A major goal of the project, then, is to explore the role of morphological structure in word recognition. We attempt to distinguish between two classes of models of morphological analysis in word recognition. In one, analogous to the dual route model of word recognition tested in Project by Fowler, morphologically complex works are decomposed in the course of lexical access to the extent that they are unfamiliar and to the extent that their morphological composition is transparent. In the second model, morphological analysis is an emergent consequence of the activation dynamics that establish mutually consistent representations of word form and word meaning. Three series of experiments are proposed that investigate the role of morphological analysis in word recognition. One series is designed to locate morphological analysis, if possible, in orthographic, phonological or semantic processing of words. A second series is designed to determine the variables that affect the likelihood that morphologically complex works undergo analysis before lexical access. The third charts the time course of morphological analysis in relation to the time courses of orthographic, phonological and semantic processing of words. The understanding of morphological analysis that we obtain from these experiments will guide our development of a network model of word recognition, and final experiments are designed to distinguish that model from the dual route model.